Reasearch And Application Of Energy Efficiency Analysis And Prediction Methods For Ethylene

Energy conservation and prediction of complex chemical processes play a crucial role in sustainable development processes.In order to analyse the influence of technology,management level and production structure on energy efficiency and energy saving potential,this paper proposes a new integrated framework that combines Index Decomposition Analysis(IDA)and energy saving potential.At the same time,this paper combines energy analysis with energy efficiency prediction.Because the data in the ethylene production process has the characteristics of high data dimensions,and its measurement process usually has some error noise and external interference,we propose a modeling method that based on Extreme Learning Machine(ELM).ELM has the ability of parallel distribution processing and a high degree of nonlinear approximation,and is not easily trapped into local minimum values.The IDA method can obtain energy activity levels,energy levels,and energy intensity based to reflect the impact of energy use.The energy-saving potential method can verify the correctness of the improvement direction proposed by the IDA method.The proposed framework can intuitively find ways to improve energy efficiency,reduce energy consumption,and save energy.Because ELM neural network can easily receive interference from noise data,we improved it by Affinity Propagation(AP)clustering method to make it more applicable to chemical processes.At the same time,the IDA+energy-saving potential hybrid framework can effectively extract feature data and remove noise,interference,and errors.Therefore,we use the IDA+energy-saving potential hybrid framework's index analysis results as inputs of the AP-ELM neural network to model the chemical devices.The research content of this paper is as follows:1.We use IDA index analysis methods to analyse the chemical data,and IDA method also has many shortcomings.Therefore,it is necessary to further distinguish the technical energy-saving potential and structural energy-saving potential of complex chemical processes,and analyse the main factors that affect their changes.Therefore,we use the hybrid framework combining IDA index analysis algorithm and energy-saving potential analysis algorithm for energy efficiency analysis.2.For the characteristics of high dimensionality,redundant information,and strong coupling of industrial data,the paper analyses and contrasts different neural network models,selects the ELM neural network with fast training speed and global approximation performance as the basic model,and combined with a new clustering method,Affinity Propagation(AP)clustering,improves the accuracy of the ELM prediction method.3.Using the hybrid framework combining IDA index analysis algorithm and energy-saving potential analysis algorithm and AP-ELM neural network model to the energy efficiency analysis and prediction process in the ethylene production process,and verify the practical application value of the two models.4.Design and build a Web-based prototype system based on B/S architecture the prototype system of ethylene energy efficiency analysis and prediction.Device data is used as a research object to verify the validity and feasibility of the proposed method.